Quantum Advantage: A Comparison of Classical and Quantum Computing for Simulating Ising Models
Category Technology Saturday - February 17 2024, 22:43 UTC - 9 months ago Recent research has shown that classical computers using a tensor network approach can outperform low error quantum computers in simulating Ising models, shedding light on the potential advantages and limitations of quantum computing. Quantum entanglement and matching the geometry of the lattice are important factors in achieving accurate simulations.
The race for quantum advantage, or the point at which quantum computers are faster than classical computers for certain tasks, has been a highly anticipated development in the world of computing. It was previously thought that with 50-100 logical low error qubits, quantum computers would be able to surpass regular supercomputers. However, new research is showing that even with 127 qubits, regular computers could beat low error quantum computers. In this article, we will explore a recent comparison of classical and quantum computing for simulating Ising models, shedding light on the potential advantages and limitations of quantum computing.
The Ising model is a mathematical model used to study the behavior of interacting particles. It has been widely used in various fields of physics and statistical mechanics, making it an ideal test case for comparing classical and quantum computing. In a recent experiment, researchers at IBM utilized their 127-qubit quantum computer to simulate a kicked Ising quantum system on the heavy hexagon lattice. This was accomplished using noise-mitigation techniques to enhance accuracy, but results still fell short of expectations.
In response, a team of physicists led by Nathan Tindall from Massey University in New Zealand developed a classical approach using tensor networks to simulate the same Ising-model problem. Tensor networks are series of data arrays connected through links, and have been used in many-body quantum systems to compress large amounts of information into a manageable size. This 'zip file' for wave functions made it possible for Tindall and his team to simulate the 127 qubits on a classical computer with an accuracy significantly higher than that of the quantum processor.
The belief-propagation (BP) method used by Tindall and his colleagues is specifically designed to reflect the geometry of the lattice, making it a better match for the IBM quantum computer than the method used by IBM researchers. By assuming the neglect of certain information about quantum entanglement between qubits, Tindall's team was able to reduce the amount of data needed by over a billion numbers, making their simulation more efficient and accurate than IBM's.
Additionally, the tensor network approach has the potential for even broader applications, including the simulation of systems with treelike correlations. Tindall and his team have shown that their method allows for simulations of the Ising-model problem to be performed to long times in the thermodynamic limit, equivalent to an infinite number of qubits, something not currently possible with quantum computers.
As the race for quantum advantage continues, this comparison of classical and quantum computing for simulating Ising models sheds light on the capabilities and limitations of quantum computers. With the recent advancements and innovations in classical methods, it is possible that classical computers will maintain a competitive edge for certain tasks. Only time will tell which approach will ultimately lead us to quantum advantage.
Share